An inertial measurement unit (IMU) typically comprises three axes of angular motion sensing and three axes of acceleration sensing. Angular motion is measured by gyroscopes and acceleration is measured by linear accelerometers. The IMUs are employed for a variety of applications including stabilization, flight guidance, and navigation, for example. Requirements on IMUs can be demanding. The units typically need to be able to accurately sense very small quantities while operating in environments that range from quiescent to highly dynamic including high rotation rates, high accelerations, as well as high levels of vibration and shock. The IMUs, and particularly the gyroscopes and accelerometers are typically designed to operate through this full range of environments. However, this involves a number of design compromises which do not adequately address optimal operations over the range of environments.
A typical gyroscope control system for a single axis fiber optic gyroscope (FOG) configuration can include a primary angular rate rebalance loop as well as auxiliary control loops comprising a phase modulator scale factor control loop, an automatic gain control loop, an offset control loop, and an intensity control loop, for example. All of the control loops operate concurrently to ensure proper operation of the FOG as an angular rate sensor. The FOG mechanization includes both analog and digital elements. In particular, the optical signal from the FOG is amplified and converted via an A/D converter to digital form, allowing the control loops to operate in the digital domain. Certain control loop parameters are chosen for best operation. For example, the analog gain is high enough to permit the naturally occurring quiescent noise in the optics and analog electronics to toggle the least significant bit of the A/D converter, so that its resolution limitation can be overcome. On the other hand, this limits the maximum signal that the A/D converter can handle before it saturates. When high dynamics are present (e.g., high vibration or shock), the A/D may saturate, leading to errors in the gyroscope readout. Also, the auxiliary loops may respond to high dynamic signals and deviate from their optimal control points.